Year: 2023
Type: dissertation
Abstract: Introduction: in anomia, electrophysiological biomarker iden0fica0on during a naming task could lead to braincomputer interfaces use in rehabilita0on (i.e., by suitable feedback or by guiding s0mula0on technics real-0me). To iden0fy such a biomarker in real-0me, it is important to explore which signal analysis is the most appropriate. The first step is to analyze naming task ERP average to seek for electrophysiological component underlying an incorrect produc0on. Our study aimed to search for amplitude and topographic differences between correct and incorrect produc0on. Methods: we studied 10 EEG of post-stroke aphasic pa0ent in a naming task, recording during 2 inclusion periods. We balanced the correct and incorrect (C1 and C2) epochs of these 2 periods. We analyzed amplitude differences by TFCE (threshold-free cluster enhancement) analysis, then topographic modula0on with tANOVA and Kruskal- Wallis tests, from de maps fiZng, in terms of global explained variance et mean dura0on. Results: we did not find amplitude or topographic sta0s0cally significant difference between C1 and C2 in the 2 periods. Conclusion: based on our results, we cannot conclude to an ERP difference between correct and incorrect produc0ons. Further inves0ga0ons might be interes0ng to do with standardized methods (e.g., more epochs analyzed, more pa0ent included with intra-pa0ent analysis) to see the reproducibility of our results. Timefrequency analyses could also be a lead to find electrophysiological biomarkers.
Source: HAL
Author: Berariu, Ilinka, Mara
Institution: Université de Bordeaux
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7008022257